Multilingual Deep Neural Network Training Using Cyclical Learning Rate.

Andreas Søeborg Kirkedal Yeon-Jun Kim

Published in: INTERSPEECH (2018)

Keyphrases

learning rate
neural network training
training algorithm
hidden layer
convergence rate
learning algorithm
convergence speed
neural network
adaptive learning rate
convergence theorem
optimization method
weight vector
multilayer neural networks
back propagation
rapid convergence
control parameters
multi objective